FusionNet: A Frame Interpolation Network for 4D Heart Models

2026-03-10

MTSAIL/LEAF/AI4Treat/MMMI/REMIA@MICCAI

AI Summary

This study addresses the challenge of insufficient temporal resolution in cardiac magnetic resonance (CMR) imaging due to short scan durations, which hampers accurate cardiac motion modeling and clinical diagnosis. To overcome this limitation, the authors propose FusionNet, a novel frame interpolation neural network that, for the first time, applies deep learning–driven frame interpolation to 4D cardiac modeling. By effectively fusing 3D cardiac shapes from adjacent time points, FusionNet generates high-fidelity, high-temporal-resolution 4D cardiac motion sequences while preserving the original short acquisition time. Experimental results demonstrate that the method achieves a Dice coefficient exceeding 0.897, outperforming existing approaches in morphological reconstruction accuracy and significantly enhancing the diagnostic utility of low-temporal-resolution CMR data.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Cardiac magnetic resonance (CMR) imaging is widely used to visualise cardiac motion and diagnose heart disease. However, standard CMR imaging requires patients to lie still in a confined space inside a loud machine for 40-60 min, which increases patient discomfort. In addition, shorter scan times decrease either or both the temporal and spatial resolutions of cardiac motion, and thus, the diagnostic accuracy of the procedure. Of these, we focus on reduced temporal resolution and propose a neural network called FusionNet to obtain four-dimensional (4D) cardiac motion with high temporal resolution from CMR images captured in a short period of time. The model estimates intermediate 3D heart shapes based on adjacent shapes. The results of an experimental evaluation of the proposed FusionNet model showed that it achieved a performance of over 0.897 in terms of the Dice coefficient, confirming that it can recover shapes more precisely than existing methods. This code is available at: https://github.com/smiyauchi199/FusionNet.git

Problem

• cardiac motion
• temporal resolution
• CMR imaging
• 4D heart models
• frame interpolation

Innovation

• FusionNet
• frame interpolation
• 4D heart modeling
• cardiac MRI
• temporal super-resolution